Summary

  • AFRINIC’s Board met online at 09:08 UTC on 20 January 2021, reviewed an undisclosed list of volunteers and appointed three community representatives and one Board representative to the 2021 Nomination Committee under Resolution 202101.597.
  • Article 9.1 excluded a Board candidate and a person domiciled in the region whose seat was open for renewal from appointment to the committee. The resolution used the unexplained phrase “South Africa sub-regions”, while the Bylaws elsewhere used “Southern Africa”. That restriction concerned NomCom membership, not who could stand for the Board.
  • The most persuasive defence is conflict prevention: a region with a seat in contest need not also supply a selector for that contest. Yet the benefit can be assessed only if AFRINIC reveals, in privacy-safe form, the regional definition, domicile test, size and treatment of the volunteer pool, appointment criteria, conflicts, reasons and correction route.
  • The committee’s power was private and procedural. AFRINIC remained a technical bookkeeper and coordinator, not a sovereign or public electoral authority; clearer evidence would make regional rotation more credible without enlarging that limited role.

L3 — The rule before the ballot

A meeting, a volunteer list and four names

The consequential part of AFRINIC’s 20 January 2021 Board meeting did not begin with a ballot paper. It began online at 09:08 UTC, with a conflict already declared. The minutes identify Vika Mpisane as the Southern Africa Board member and record his conflict of interest for item 7, Nomination Committee Appointments. When that item arrived, the Board reviewed the list received after a public call for volunteers. The public record does not reveal how long the list was or name everybody on it. It then records the appointment.

Resolution 202101.597 said that four people were required for the 2021 Nomination Committee under Article 9 of the 2016 Bylaws: three community representatives and one Board member. Caleb Ogundele Olumuyiwa, Fabian Arbogast and Raymond Mamattah were appointed as the three community representatives. Abdalla Omari was appointed as the Board representative. Benjamin Eshun proposed the resolution and Adewale Adedokun seconded it. Mpisane recused himself, Eddy Kayihura abstained, and the resolution passed.

Those details matter because they distinguish an actual institutional act from an abstract concern about elections. There was a governing bylaw, a volunteer call, an internal review, a named resolution, four named appointees and recorded voting states. There was also a geographic recital. The resolution said that people domiciled in “South Africa sub-regions” were not eligible under Article 9.1. The effect of that recital was to narrow the pool from which the Board made the four-person appointment.

The wording needs disciplined treatment. Article 9.1 did not say that South African candidates, Southern African candidates or candidates from any particular country could not stand for the Board. It disqualified a person from appointment to NomCom if that person was a Board candidate or was domiciled in a region whose seat was open for renewal during an election. The restriction attached to the selector role. Its object was the committee, not the Board candidacy itself.

Nor can three different formulations be silently collapsed into one. Resolution 202101.597 used “South Africa sub-regions”. Article 9.1 referred more generally to “a region whose seat is open for renewal”. Elsewhere, Article 13.5 of the same Bylaws used “Southern Africa”. The preserved public material does not reconcile those terms. “South Africa” commonly looks like a country name; “Southern Africa” looks like a regional designation; “sub-regions” is plural. The record supports recognition of the discrepancy, not an editorial repair of it.

Treating the phrases as interchangeable would supply precision that the instrument itself did not publicly supply.

That restraint is central to any fair reading of the resolution. The appointment may have been perfectly consistent with an internal understanding available to those present. Private material may have identified the open seat and supplied a settled regional interpretation. The public evidence simply does not show it. Absence from the public record is not proof that no underlying document existed. It does mean a member or operator relying on the published record cannot reproduce the classification from the published record alone.

What Article 9 made the committee capable of doing

The geographic filter becomes significant only when connected to the committee’s assigned function. Article 9 established a Nomination Committee consisting of a chairman and three other members appointed by the Board, following a public call for voluntary nominations from the African Internet Community. Article 9.2 required the committee to report to the Board and work under guidelines prescribed by the Board. Article 9.3 then gave it a set of functions that reached directly into the candidate field.

The committee could seek a satisfactory number of candidates, issue a call for candidates, prescribe candidate eligibility criteria, interview candidates if it chose and finalise the candidate list. It also had a role in supervising polling conducted by the Election Committee. These were not merely clerical tasks. The capacity to define eligibility, interview nominees and finalise a list placed NomCom upstream of the member vote. It could shape the field from which members would later choose.

The resulting sequence is simple but easy to obscure. First, Article 9.1 defined who could not be appointed to NomCom. Secondly, the Board reviewed volunteers and chose four committee members from the remaining eligible field. Thirdly, Article 9.3 equipped that committee to influence which Board candidates entered the final list. Only after those stages did members encounter a shaped choice. The geography rule therefore operated before the ballot even though it did not disqualify Board candidates on geographic grounds.

This distinction also explains why formal voting rights are not the whole governance story. A member can retain an undiminished right to cast a vote while facing a field shaped by prior decisions about selectors, criteria and interviews. That is not necessarily improper; virtually every organised nomination process contains some upstream structure. But the structure deserves the same clarity as the eventual count because it affects the range and character of the available choice.

The 2013 AFRINIC appointment examined by BTW provides a useful institutional comparator, but only a comparator. It shows why a four-person nomination gate can matter before voting and why criteria, conflicts and reasons deserve public attention. It does not prove anything about the conduct, motives or validity of the 2021 Board or its appointees. The two appointments are separate acts. The relevant lesson is structural: when a small appointed group can influence access to a later ballot, the appointment record is part of election accountability even though the group is not itself the electorate.

The strongest case for geographic distance

The most serious analysis must begin with the best defence of the restriction. A region whose Board seat is being renewed has an obvious interest in the contest. If that region also supplies someone to the committee that defines eligibility, interviews nominees and finalises the candidate list, observers may reasonably worry that the same regional constituency has influence on both the supply of candidates and the gate through which they pass. A domicile exclusion creates distance between the contest and its selectors.

Regional rotation can also distribute institutional service. AFRINIC’s operating area is broad, and a rule that prevents one region from occupying both the open-seat position and a place on NomCom can widen participation over time. The safeguard may encourage people elsewhere to volunteer, make the committee visibly less concentrated and protect confidence in a Board whose structure recognises regions. A rule need not prove actual bias before acting against an avoidable appearance of conflict. Preventive design often works precisely by separating roles before motives need to be judged.

The minutes contain further positive signals. Mpisane declared a conflict in advance of item 7 and recused himself. Kayihura abstained. Those are distinct acts, and the public record does not reveal the precise basis of the conflict or the reason for the abstention. Even so, recording them is better than presenting the appointment as unanimous and frictionless. Recusal removes a conflicted participant from the relevant decision. Abstention tells readers that a director did not cast an affirmative or negative vote. The entries preserve procedural states that might otherwise disappear into the fact that the resolution passed.

No available evidence proves that Olumuyiwa, Arbogast, Mamattah or Omari acted improperly. It does not prove that the Board manipulated the volunteer field, chose unsuitable people, made a defective appointment or produced an invalid election. A geographic eligibility rule is not itself proof of discrimination in an illicit sense. Within a private association organised partly around regions, it can be a proportionate internal conflict safeguard.

The benign case is strongest when expressed narrowly. The rule could reduce direct conflict risk in the nomination stage, disperse opportunity to serve and protect confidence in regional rotation. The declared conflict, recusal and abstention show that the Board recognised procedural restraint as relevant. None of this requires portraying NomCom as a constitutional body or the regional categories as political representation of African peoples. It is enough to say that a private organisation may separate internal roles to protect the credibility of its own election arrangements.

The cost paid before anyone votes

The safeguard nevertheless has a price. Domicile can be a rough proxy for conflict. Excluding everybody within a whole domicile category removes people regardless of their individual independence, expertise or distance from a candidate. It can take local technical and institutional knowledge out of the room precisely when nominees from the region and the circumstances of its open seat may need careful understanding. The prophylactic rule treats geographic connection as more important than a person-specific assessment for this appointment purpose.

It can also shrink an already limited volunteer field. The minutes say the Board reviewed a list of volunteers, but they do not state the denominator. If there were many eligible volunteers across several regions, the exclusion may have had little practical effect on committee quality or diversity. If there were few, removing one domicile category may have materially constrained the available expertise. Both possibilities fit the published account. Without privacy-safe aggregate numbers, a reader cannot know which describes the 2021 choice.

This is why the phrase in the resolution is more than a drafting curiosity. Eligibility classifications require boundaries. Was “South Africa sub-regions” intended to mean the country of South Africa, the Bylaws’ Southern Africa region, more than one sub-region, or something else understood internally? Which seat was open for renewal and linked to that wording? What counted as domicile: permanent residence, principal home, legal residence, ordinary residence, employer location or another measure? The public material answers none of those questions.

Borderline cases deepen the issue. A person may recently have moved, maintain homes in more than one place or have professional activity in a region different from their residence. Domicile can carry a technical legal meaning in some settings and an ordinary factual meaning in others. The record does not reveal what evidence the Board expected, how it treated dual or changing circumstances, who made a disputed classification or whether an affected volunteer could correct an error. It would be unsound to invent answers. It is equally unsound to assume the classification was self-executing merely because the Bylaws named a category.

The public material also leaves the comparative decision opaque. It does not disclose the complete volunteer population, individual eligibility outcomes, appointment criteria, assessments, interviews, rejection reasons or reasons for preferring the four appointees. It does not explain Mpisane’s precise conflict or Kayihura’s abstention. It identifies no review or correction route. These are gaps in what the public can verify, not evidence that the Board possessed no criteria, held no discussion or kept no internal record.

The proper finding at this level is therefore bounded. Resolution 202101.597 visibly applied a geographic eligibility recital before appointing a four-person committee with later influence over the candidate field. Regional distance could serve a legitimate conflict-prevention purpose. The same exclusion could remove useful knowledge and reduce the selector pool. The published record is too thin to measure either effect because it omits the definition, denominator, test, comparisons and correction mechanism that would make the balance reproducible.